Update README.md
Browse files
README.md
CHANGED
@@ -1,241 +1,241 @@
|
|
1 |
-
---
|
2 |
-
tags:
|
3 |
-
- setfit
|
4 |
-
- sentence-transformers
|
5 |
-
- text-classification
|
6 |
-
- generated_from_setfit_trainer
|
7 |
-
widget:
|
8 |
-
- text: Mujhe apne galtiyon ka ehsaas hai aur main unke liye maafi chahta hoon.
|
9 |
-
- text: Mujhe yeh step samajhne mein dikkat ho rahi hai, kya aap madad kar sakte hain?
|
10 |
-
- text: Mujhe abhi tak kuch update kyun nahi mila, yeh bahut frustrating hai.
|
11 |
-
- text: Is app ka loading time mujhe thoda zyada lagta hai.
|
12 |
-
- text: Kya aap mujhe is event ki timing bata sakte hain?
|
13 |
-
metrics:
|
14 |
-
- accuracy
|
15 |
-
pipeline_tag: text-classification
|
16 |
-
library_name: setfit
|
17 |
-
inference: true
|
18 |
-
base_model: MoritzLaurer/mDeBERTa-v3-base-mnli-xnli
|
19 |
-
model-index:
|
20 |
-
- name: SetFit with MoritzLaurer/mDeBERTa-v3-base-mnli-xnli
|
21 |
-
results:
|
22 |
-
- task:
|
23 |
-
type: text-classification
|
24 |
-
name: Text Classification
|
25 |
-
dataset:
|
26 |
-
name: Unknown
|
27 |
-
type: unknown
|
28 |
-
split: test
|
29 |
-
metrics:
|
30 |
-
- type: accuracy
|
31 |
-
value: 0.32
|
32 |
-
name: Accuracy
|
33 |
-
---
|
34 |
-
|
35 |
-
# SetFit with MoritzLaurer/mDeBERTa-v3-base-mnli-xnli
|
36 |
-
|
37 |
-
This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [MoritzLaurer/mDeBERTa-v3-base-mnli-xnli](https://huggingface.co/MoritzLaurer/mDeBERTa-v3-base-mnli-xnli) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance is used for classification.
|
38 |
-
|
39 |
-
The model has been trained using an efficient few-shot learning technique that involves:
|
40 |
-
|
41 |
-
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
|
42 |
-
2. Training a classification head with features from the fine-tuned Sentence Transformer.
|
43 |
-
|
44 |
-
## Model Details
|
45 |
-
|
46 |
-
### Model Description
|
47 |
-
- **Model Type:** SetFit
|
48 |
-
- **Sentence Transformer body:** [MoritzLaurer/mDeBERTa-v3-base-mnli-xnli](https://huggingface.co/MoritzLaurer/mDeBERTa-v3-base-mnli-xnli)
|
49 |
-
- **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance
|
50 |
-
- **Maximum Sequence Length:** 512 tokens
|
51 |
-
- **Number of Classes:** 19 classes
|
52 |
-
<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
|
53 |
-
<!-- - **Language:** Unknown -->
|
54 |
-
<!-- - **License:** Unknown -->
|
55 |
-
|
56 |
-
### Model Sources
|
57 |
-
|
58 |
-
- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
|
59 |
-
- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
|
60 |
-
- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
|
61 |
-
|
62 |
-
### Model Labels
|
63 |
-
| Label | Examples |
|
64 |
-
|:------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
|
65 |
-
| 4 | <ul><li>'Yeh rahin wo steps jisse aap apni payment kar sakte hain.'</li><li>'Kya aap mujhe yeh batane ka tarika thoda aasan kar sakte hain?'</li><li>'Is option ke madhyam se aap apni queries kaise solve kar sakte hain, jaan lijiye.'</li></ul> |
|
66 |
-
| 16 | <ul><li>'Aapke feedback ko humne dhyan mein rakha hai.'</li><li>'Yeh galti humare systems ki wajah se hui hai.'</li><li>'Mujhe is samasya ko suljhane mein zyada samay lena nahi chahiye tha.'</li></ul> |
|
67 |
-
| 8 | <ul><li>'Main aapko pareshan karne ke liye maafi chahta hoon.'</li><li>'Humein is samasya ke liye maafi chahiye.'</li><li>'Mere kaam se agar aapko takleef hui ho, toh mujhe maaf kar dijiye.'</li></ul> |
|
68 |
-
| 13 | <ul><li>'Mujhe yeh clarify karne ki zarurat hai ki agla step kya hai?'</li><li>'Mujhe pata karna hai ki maine jo complaint ki thi uska kya hua.'</li><li>'Mujhe bataye ki pehle kitne payments honge iss plan ke liye.'</li></ul> |
|
69 |
-
| 15 | <ul><li>'Yeh features sahi hai, lekin kuch aur additional functionalities honi chahiye.'</li><li>'Product ke size ki jankari hamesha saaf honi chahiye.'</li><li>'Main chahunga ki online form aur simple ho.'</li></ul> |
|
70 |
-
| 12 | <ul><li>'Mujhe product ke sath kuch samasya hai.'</li><li>'Mera phone charging nahi ho raha.'</li><li>'Mujhe courier service mein dikkat hai, report karna hai.'</li></ul> |
|
71 |
-
| 11 | <ul><li>'Mujhe samajh nahi aa raha, is offer mein koi chhupi shartein toh nahi hai?'</li><li>'Kis tarah se main feedback de sakta hoon?'</li><li>'Kya koi referral program hai jo mujhe join karna chahiye?'</li></ul> |
|
72 |
-
| 2 | <ul><li>'Item ke sath saathi accessories nahi mil rahe hain.'</li><li>'Aap logon ne jo samay liya, wo bilkul zyada tha.'</li><li>'Meri order delivery mein bahut der ho gayi hai.'</li></ul> |
|
73 |
-
| 18 | <ul><li>'Mujhe yeh bilkul pasand nahi hai ki meri baat ignore ki gayi.'</li><li>'Kam ke liye mera dosto ka support bahut sukhdayak hai.'</li><li>'Aaj ka din kaafi udaas beete raha hai.'</li></ul> |
|
74 |
-
| 14 | <ul><li>'Kya main kal ki delivery ko agle hafte reschedule kar sakta/sakti hoon?'</li><li>'Mujhe refund ke liye kya documents chahiye?'</li><li>'Kya main appointment ko dobara set kar sakta/sakti hoon?'</li></ul> |
|
75 |
-
| 7 | <ul><li>'Main aapko dhanyavad dena chahta hoon, aapne meri madad ki.'</li><li>'Aapne jo kiya, uske liye aapko sabse pehle prashansha milni chahiye.'</li><li>'Aapka samay dene ke liye abhaar.'</li></ul> |
|
76 |
-
| 3 | <ul><li>'Mujhe kisi event ke tickets ka status check karna hai.'</li><li>'Kya aap mujhe customer support number de sakte hain?'</li><li>'Main apne account ka balance kaise check kar sakta/sakti hoon?'</li></ul> |
|
77 |
-
| 5 | <ul><li>'Alvida, tumhara din acha rahe!'</li><li>'Hello! Aaj aap kaise hain?'</li><li>'Swagat hai! Kya main aapki kuch madad kar sakta hoon?'</li></ul> |
|
78 |
-
| 0 | <ul><li>'Mujhe kuch samajh nahi aa raha hai, kya mujhe thoda aur samjha sakte hain?'</li><li>'Agar main aisa karoon, to kya kuch badal jaayega? Main sure nahi hoon.'</li><li>'Yeh product ki warranty ki details clear nahi hain.'</li></ul> |
|
79 |
-
| 6 | <ul><li>'Chalo, alvida bolte hain!'</li><li>'Phir se baat karte hain!'</li><li>'Adieu, aapka din shubh ho!'</li></ul> |
|
80 |
-
| 17 | <ul><li>'Mere account mein login karne mein dikkat aa rahi hai, madad karein.'</li><li>'Mujhe apne account mein login karne mein madad chahiye.'</li><li>'Kya aap mujhe terms and conditions ke details de sakte hain?'</li></ul> |
|
81 |
-
| 10 | <ul><li>'Main aapki baat se sehmat hoon.'</li><li>'Mujhe yeh batayein ki meri booking sahi hai na?'</li></ul> |
|
82 |
-
| 9 | <ul><li>'Kya aap mujhe yeh concept aur clear kar sakte hain?'</li><li>'Mujhe yeh samajhne mein dikkat ho rahi hai, kya aap vyakhya de sakte hain?'</li></ul> |
|
83 |
-
| 1 | <ul><li>'Aaj dosto ke sath waqt bitana bahut acha laga.'</li><li>'Aaj baarish me bheegna bahut refreshing tha, mujhe yeh moment pasand aaya.'</li><li>'Aapka support bahut madadgar raha.'</li></ul> |
|
84 |
-
|
85 |
-
## Evaluation
|
86 |
-
|
87 |
-
### Metrics
|
88 |
-
| Label | Accuracy |
|
89 |
-
|:--------|:---------|
|
90 |
-
| **all** | 0.32 |
|
91 |
-
|
92 |
-
## Uses
|
93 |
-
|
94 |
-
### Direct Use for Inference
|
95 |
-
|
96 |
-
First install the SetFit library:
|
97 |
-
|
98 |
-
```bash
|
99 |
-
pip install setfit
|
100 |
-
```
|
101 |
-
|
102 |
-
Then you can load this model and run inference.
|
103 |
-
|
104 |
-
```python
|
105 |
-
from setfit import SetFitModel
|
106 |
-
|
107 |
-
# Download from the 🤗 Hub
|
108 |
-
model = SetFitModel.from_pretrained("
|
109 |
-
# Run inference
|
110 |
-
preds = model("Kya aap mujhe is event ki timing bata sakte hain?")
|
111 |
-
```
|
112 |
-
|
113 |
-
<!--
|
114 |
-
### Downstream Use
|
115 |
-
|
116 |
-
*List how someone could finetune this model on their own dataset.*
|
117 |
-
-->
|
118 |
-
|
119 |
-
<!--
|
120 |
-
### Out-of-Scope Use
|
121 |
-
|
122 |
-
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
|
123 |
-
-->
|
124 |
-
|
125 |
-
<!--
|
126 |
-
## Bias, Risks and Limitations
|
127 |
-
|
128 |
-
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
|
129 |
-
-->
|
130 |
-
|
131 |
-
<!--
|
132 |
-
### Recommendations
|
133 |
-
|
134 |
-
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
|
135 |
-
-->
|
136 |
-
|
137 |
-
## Training Details
|
138 |
-
|
139 |
-
### Training Set Metrics
|
140 |
-
| Training set | Min | Median | Max |
|
141 |
-
|:-------------|:----|:-------|:----|
|
142 |
-
| Word count | 3 | 9.76 | 15 |
|
143 |
-
|
144 |
-
| Label | Training Sample Count |
|
145 |
-
|:------|:----------------------|
|
146 |
-
| 0 | 6 |
|
147 |
-
| 1 | 3 |
|
148 |
-
| 2 | 3 |
|
149 |
-
| 3 | 5 |
|
150 |
-
| 4 | 7 |
|
151 |
-
| 5 | 3 |
|
152 |
-
| 6 | 6 |
|
153 |
-
| 7 | 8 |
|
154 |
-
| 8 | 6 |
|
155 |
-
| 9 | 2 |
|
156 |
-
| 10 | 2 |
|
157 |
-
| 11 | 5 |
|
158 |
-
| 12 | 6 |
|
159 |
-
| 13 | 5 |
|
160 |
-
| 14 | 9 |
|
161 |
-
| 15 | 9 |
|
162 |
-
| 16 | 9 |
|
163 |
-
| 17 | 3 |
|
164 |
-
| 18 | 3 |
|
165 |
-
|
166 |
-
### Training Hyperparameters
|
167 |
-
- batch_size: (16, 2)
|
168 |
-
- num_epochs: (1, 16)
|
169 |
-
- max_steps: -1
|
170 |
-
- sampling_strategy: oversampling
|
171 |
-
- body_learning_rate: (2e-05, 1e-05)
|
172 |
-
- head_learning_rate: 0.01
|
173 |
-
- loss: CosineSimilarityLoss
|
174 |
-
- distance_metric: cosine_distance
|
175 |
-
- margin: 0.25
|
176 |
-
- end_to_end: False
|
177 |
-
- use_amp: False
|
178 |
-
- warmup_proportion: 0.1
|
179 |
-
- l2_weight: 0.01
|
180 |
-
- seed: 42
|
181 |
-
- eval_max_steps: -1
|
182 |
-
- load_best_model_at_end: False
|
183 |
-
|
184 |
-
### Training Results
|
185 |
-
| Epoch | Step | Training Loss | Validation Loss |
|
186 |
-
|:------:|:----:|:-------------:|:---------------:|
|
187 |
-
| 0.0017 | 1 | 0.2335 | - |
|
188 |
-
| 0.0853 | 50 | 0.2514 | - |
|
189 |
-
| 0.1706 | 100 | 0.1619 | - |
|
190 |
-
| 0.2560 | 150 | 0.1124 | - |
|
191 |
-
| 0.3413 | 200 | 0.078 | - |
|
192 |
-
| 0.4266 | 250 | 0.0623 | - |
|
193 |
-
| 0.5119 | 300 | 0.0576 | - |
|
194 |
-
| 0.5973 | 350 | 0.0421 | - |
|
195 |
-
| 0.6826 | 400 | 0.0391 | - |
|
196 |
-
| 0.7679 | 450 | 0.0386 | - |
|
197 |
-
| 0.8532 | 500 | 0.0302 | - |
|
198 |
-
| 0.9386 | 550 | 0.0245 | - |
|
199 |
-
|
200 |
-
### Framework Versions
|
201 |
-
- Python: 3.10.16
|
202 |
-
- SetFit: 1.1.1
|
203 |
-
- Sentence Transformers: 3.3.1
|
204 |
-
- Transformers: 4.46.3
|
205 |
-
- PyTorch: 2.5.1+cpu
|
206 |
-
- Datasets: 3.2.0
|
207 |
-
- Tokenizers: 0.20.3
|
208 |
-
|
209 |
-
## Citation
|
210 |
-
|
211 |
-
### BibTeX
|
212 |
-
```bibtex
|
213 |
-
@article{https://doi.org/10.48550/arxiv.2209.11055,
|
214 |
-
doi = {10.48550/ARXIV.2209.11055},
|
215 |
-
url = {https://arxiv.org/abs/2209.11055},
|
216 |
-
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
|
217 |
-
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
|
218 |
-
title = {Efficient Few-Shot Learning Without Prompts},
|
219 |
-
publisher = {arXiv},
|
220 |
-
year = {2022},
|
221 |
-
copyright = {Creative Commons Attribution 4.0 International}
|
222 |
-
}
|
223 |
-
```
|
224 |
-
|
225 |
-
<!--
|
226 |
-
## Glossary
|
227 |
-
|
228 |
-
*Clearly define terms in order to be accessible across audiences.*
|
229 |
-
-->
|
230 |
-
|
231 |
-
<!--
|
232 |
-
## Model Card Authors
|
233 |
-
|
234 |
-
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
|
235 |
-
-->
|
236 |
-
|
237 |
-
<!--
|
238 |
-
## Model Card Contact
|
239 |
-
|
240 |
-
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
|
241 |
-->
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- setfit
|
4 |
+
- sentence-transformers
|
5 |
+
- text-classification
|
6 |
+
- generated_from_setfit_trainer
|
7 |
+
widget:
|
8 |
+
- text: Mujhe apne galtiyon ka ehsaas hai aur main unke liye maafi chahta hoon.
|
9 |
+
- text: Mujhe yeh step samajhne mein dikkat ho rahi hai, kya aap madad kar sakte hain?
|
10 |
+
- text: Mujhe abhi tak kuch update kyun nahi mila, yeh bahut frustrating hai.
|
11 |
+
- text: Is app ka loading time mujhe thoda zyada lagta hai.
|
12 |
+
- text: Kya aap mujhe is event ki timing bata sakte hain?
|
13 |
+
metrics:
|
14 |
+
- accuracy
|
15 |
+
pipeline_tag: text-classification
|
16 |
+
library_name: setfit
|
17 |
+
inference: true
|
18 |
+
base_model: MoritzLaurer/mDeBERTa-v3-base-mnli-xnli
|
19 |
+
model-index:
|
20 |
+
- name: SetFit with MoritzLaurer/mDeBERTa-v3-base-mnli-xnli
|
21 |
+
results:
|
22 |
+
- task:
|
23 |
+
type: text-classification
|
24 |
+
name: Text Classification
|
25 |
+
dataset:
|
26 |
+
name: Unknown
|
27 |
+
type: unknown
|
28 |
+
split: test
|
29 |
+
metrics:
|
30 |
+
- type: accuracy
|
31 |
+
value: 0.32
|
32 |
+
name: Accuracy
|
33 |
+
---
|
34 |
+
|
35 |
+
# SetFit with MoritzLaurer/mDeBERTa-v3-base-mnli-xnli
|
36 |
+
|
37 |
+
This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [MoritzLaurer/mDeBERTa-v3-base-mnli-xnli](https://huggingface.co/MoritzLaurer/mDeBERTa-v3-base-mnli-xnli) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance is used for classification.
|
38 |
+
|
39 |
+
The model has been trained using an efficient few-shot learning technique that involves:
|
40 |
+
|
41 |
+
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
|
42 |
+
2. Training a classification head with features from the fine-tuned Sentence Transformer.
|
43 |
+
|
44 |
+
## Model Details
|
45 |
+
|
46 |
+
### Model Description
|
47 |
+
- **Model Type:** SetFit
|
48 |
+
- **Sentence Transformer body:** [MoritzLaurer/mDeBERTa-v3-base-mnli-xnli](https://huggingface.co/MoritzLaurer/mDeBERTa-v3-base-mnli-xnli)
|
49 |
+
- **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance
|
50 |
+
- **Maximum Sequence Length:** 512 tokens
|
51 |
+
- **Number of Classes:** 19 classes
|
52 |
+
<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
|
53 |
+
<!-- - **Language:** Unknown -->
|
54 |
+
<!-- - **License:** Unknown -->
|
55 |
+
|
56 |
+
### Model Sources
|
57 |
+
|
58 |
+
- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
|
59 |
+
- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
|
60 |
+
- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
|
61 |
+
|
62 |
+
### Model Labels
|
63 |
+
| Label | Examples |
|
64 |
+
|:------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
|
65 |
+
| 4 | <ul><li>'Yeh rahin wo steps jisse aap apni payment kar sakte hain.'</li><li>'Kya aap mujhe yeh batane ka tarika thoda aasan kar sakte hain?'</li><li>'Is option ke madhyam se aap apni queries kaise solve kar sakte hain, jaan lijiye.'</li></ul> |
|
66 |
+
| 16 | <ul><li>'Aapke feedback ko humne dhyan mein rakha hai.'</li><li>'Yeh galti humare systems ki wajah se hui hai.'</li><li>'Mujhe is samasya ko suljhane mein zyada samay lena nahi chahiye tha.'</li></ul> |
|
67 |
+
| 8 | <ul><li>'Main aapko pareshan karne ke liye maafi chahta hoon.'</li><li>'Humein is samasya ke liye maafi chahiye.'</li><li>'Mere kaam se agar aapko takleef hui ho, toh mujhe maaf kar dijiye.'</li></ul> |
|
68 |
+
| 13 | <ul><li>'Mujhe yeh clarify karne ki zarurat hai ki agla step kya hai?'</li><li>'Mujhe pata karna hai ki maine jo complaint ki thi uska kya hua.'</li><li>'Mujhe bataye ki pehle kitne payments honge iss plan ke liye.'</li></ul> |
|
69 |
+
| 15 | <ul><li>'Yeh features sahi hai, lekin kuch aur additional functionalities honi chahiye.'</li><li>'Product ke size ki jankari hamesha saaf honi chahiye.'</li><li>'Main chahunga ki online form aur simple ho.'</li></ul> |
|
70 |
+
| 12 | <ul><li>'Mujhe product ke sath kuch samasya hai.'</li><li>'Mera phone charging nahi ho raha.'</li><li>'Mujhe courier service mein dikkat hai, report karna hai.'</li></ul> |
|
71 |
+
| 11 | <ul><li>'Mujhe samajh nahi aa raha, is offer mein koi chhupi shartein toh nahi hai?'</li><li>'Kis tarah se main feedback de sakta hoon?'</li><li>'Kya koi referral program hai jo mujhe join karna chahiye?'</li></ul> |
|
72 |
+
| 2 | <ul><li>'Item ke sath saathi accessories nahi mil rahe hain.'</li><li>'Aap logon ne jo samay liya, wo bilkul zyada tha.'</li><li>'Meri order delivery mein bahut der ho gayi hai.'</li></ul> |
|
73 |
+
| 18 | <ul><li>'Mujhe yeh bilkul pasand nahi hai ki meri baat ignore ki gayi.'</li><li>'Kam ke liye mera dosto ka support bahut sukhdayak hai.'</li><li>'Aaj ka din kaafi udaas beete raha hai.'</li></ul> |
|
74 |
+
| 14 | <ul><li>'Kya main kal ki delivery ko agle hafte reschedule kar sakta/sakti hoon?'</li><li>'Mujhe refund ke liye kya documents chahiye?'</li><li>'Kya main appointment ko dobara set kar sakta/sakti hoon?'</li></ul> |
|
75 |
+
| 7 | <ul><li>'Main aapko dhanyavad dena chahta hoon, aapne meri madad ki.'</li><li>'Aapne jo kiya, uske liye aapko sabse pehle prashansha milni chahiye.'</li><li>'Aapka samay dene ke liye abhaar.'</li></ul> |
|
76 |
+
| 3 | <ul><li>'Mujhe kisi event ke tickets ka status check karna hai.'</li><li>'Kya aap mujhe customer support number de sakte hain?'</li><li>'Main apne account ka balance kaise check kar sakta/sakti hoon?'</li></ul> |
|
77 |
+
| 5 | <ul><li>'Alvida, tumhara din acha rahe!'</li><li>'Hello! Aaj aap kaise hain?'</li><li>'Swagat hai! Kya main aapki kuch madad kar sakta hoon?'</li></ul> |
|
78 |
+
| 0 | <ul><li>'Mujhe kuch samajh nahi aa raha hai, kya mujhe thoda aur samjha sakte hain?'</li><li>'Agar main aisa karoon, to kya kuch badal jaayega? Main sure nahi hoon.'</li><li>'Yeh product ki warranty ki details clear nahi hain.'</li></ul> |
|
79 |
+
| 6 | <ul><li>'Chalo, alvida bolte hain!'</li><li>'Phir se baat karte hain!'</li><li>'Adieu, aapka din shubh ho!'</li></ul> |
|
80 |
+
| 17 | <ul><li>'Mere account mein login karne mein dikkat aa rahi hai, madad karein.'</li><li>'Mujhe apne account mein login karne mein madad chahiye.'</li><li>'Kya aap mujhe terms and conditions ke details de sakte hain?'</li></ul> |
|
81 |
+
| 10 | <ul><li>'Main aapki baat se sehmat hoon.'</li><li>'Mujhe yeh batayein ki meri booking sahi hai na?'</li></ul> |
|
82 |
+
| 9 | <ul><li>'Kya aap mujhe yeh concept aur clear kar sakte hain?'</li><li>'Mujhe yeh samajhne mein dikkat ho rahi hai, kya aap vyakhya de sakte hain?'</li></ul> |
|
83 |
+
| 1 | <ul><li>'Aaj dosto ke sath waqt bitana bahut acha laga.'</li><li>'Aaj baarish me bheegna bahut refreshing tha, mujhe yeh moment pasand aaya.'</li><li>'Aapka support bahut madadgar raha.'</li></ul> |
|
84 |
+
|
85 |
+
## Evaluation
|
86 |
+
|
87 |
+
### Metrics
|
88 |
+
| Label | Accuracy |
|
89 |
+
|:--------|:---------|
|
90 |
+
| **all** | 0.32 |
|
91 |
+
|
92 |
+
## Uses
|
93 |
+
|
94 |
+
### Direct Use for Inference
|
95 |
+
|
96 |
+
First install the SetFit library:
|
97 |
+
|
98 |
+
```bash
|
99 |
+
pip install setfit
|
100 |
+
```
|
101 |
+
|
102 |
+
Then you can load this model and run inference.
|
103 |
+
|
104 |
+
```python
|
105 |
+
from setfit import SetFitModel
|
106 |
+
|
107 |
+
# Download from the 🤗 Hub
|
108 |
+
model = SetFitModel.from_pretrained("rbojja/FT-mDeBERTa-v3-base-mnli-xnli")
|
109 |
+
# Run inference
|
110 |
+
preds = model("Kya aap mujhe is event ki timing bata sakte hain?")
|
111 |
+
```
|
112 |
+
|
113 |
+
<!--
|
114 |
+
### Downstream Use
|
115 |
+
|
116 |
+
*List how someone could finetune this model on their own dataset.*
|
117 |
+
-->
|
118 |
+
|
119 |
+
<!--
|
120 |
+
### Out-of-Scope Use
|
121 |
+
|
122 |
+
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
|
123 |
+
-->
|
124 |
+
|
125 |
+
<!--
|
126 |
+
## Bias, Risks and Limitations
|
127 |
+
|
128 |
+
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
|
129 |
+
-->
|
130 |
+
|
131 |
+
<!--
|
132 |
+
### Recommendations
|
133 |
+
|
134 |
+
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
|
135 |
+
-->
|
136 |
+
|
137 |
+
## Training Details
|
138 |
+
|
139 |
+
### Training Set Metrics
|
140 |
+
| Training set | Min | Median | Max |
|
141 |
+
|:-------------|:----|:-------|:----|
|
142 |
+
| Word count | 3 | 9.76 | 15 |
|
143 |
+
|
144 |
+
| Label | Training Sample Count |
|
145 |
+
|:------|:----------------------|
|
146 |
+
| 0 | 6 |
|
147 |
+
| 1 | 3 |
|
148 |
+
| 2 | 3 |
|
149 |
+
| 3 | 5 |
|
150 |
+
| 4 | 7 |
|
151 |
+
| 5 | 3 |
|
152 |
+
| 6 | 6 |
|
153 |
+
| 7 | 8 |
|
154 |
+
| 8 | 6 |
|
155 |
+
| 9 | 2 |
|
156 |
+
| 10 | 2 |
|
157 |
+
| 11 | 5 |
|
158 |
+
| 12 | 6 |
|
159 |
+
| 13 | 5 |
|
160 |
+
| 14 | 9 |
|
161 |
+
| 15 | 9 |
|
162 |
+
| 16 | 9 |
|
163 |
+
| 17 | 3 |
|
164 |
+
| 18 | 3 |
|
165 |
+
|
166 |
+
### Training Hyperparameters
|
167 |
+
- batch_size: (16, 2)
|
168 |
+
- num_epochs: (1, 16)
|
169 |
+
- max_steps: -1
|
170 |
+
- sampling_strategy: oversampling
|
171 |
+
- body_learning_rate: (2e-05, 1e-05)
|
172 |
+
- head_learning_rate: 0.01
|
173 |
+
- loss: CosineSimilarityLoss
|
174 |
+
- distance_metric: cosine_distance
|
175 |
+
- margin: 0.25
|
176 |
+
- end_to_end: False
|
177 |
+
- use_amp: False
|
178 |
+
- warmup_proportion: 0.1
|
179 |
+
- l2_weight: 0.01
|
180 |
+
- seed: 42
|
181 |
+
- eval_max_steps: -1
|
182 |
+
- load_best_model_at_end: False
|
183 |
+
|
184 |
+
### Training Results
|
185 |
+
| Epoch | Step | Training Loss | Validation Loss |
|
186 |
+
|:------:|:----:|:-------------:|:---------------:|
|
187 |
+
| 0.0017 | 1 | 0.2335 | - |
|
188 |
+
| 0.0853 | 50 | 0.2514 | - |
|
189 |
+
| 0.1706 | 100 | 0.1619 | - |
|
190 |
+
| 0.2560 | 150 | 0.1124 | - |
|
191 |
+
| 0.3413 | 200 | 0.078 | - |
|
192 |
+
| 0.4266 | 250 | 0.0623 | - |
|
193 |
+
| 0.5119 | 300 | 0.0576 | - |
|
194 |
+
| 0.5973 | 350 | 0.0421 | - |
|
195 |
+
| 0.6826 | 400 | 0.0391 | - |
|
196 |
+
| 0.7679 | 450 | 0.0386 | - |
|
197 |
+
| 0.8532 | 500 | 0.0302 | - |
|
198 |
+
| 0.9386 | 550 | 0.0245 | - |
|
199 |
+
|
200 |
+
### Framework Versions
|
201 |
+
- Python: 3.10.16
|
202 |
+
- SetFit: 1.1.1
|
203 |
+
- Sentence Transformers: 3.3.1
|
204 |
+
- Transformers: 4.46.3
|
205 |
+
- PyTorch: 2.5.1+cpu
|
206 |
+
- Datasets: 3.2.0
|
207 |
+
- Tokenizers: 0.20.3
|
208 |
+
|
209 |
+
## Citation
|
210 |
+
|
211 |
+
### BibTeX
|
212 |
+
```bibtex
|
213 |
+
@article{https://doi.org/10.48550/arxiv.2209.11055,
|
214 |
+
doi = {10.48550/ARXIV.2209.11055},
|
215 |
+
url = {https://arxiv.org/abs/2209.11055},
|
216 |
+
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
|
217 |
+
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
|
218 |
+
title = {Efficient Few-Shot Learning Without Prompts},
|
219 |
+
publisher = {arXiv},
|
220 |
+
year = {2022},
|
221 |
+
copyright = {Creative Commons Attribution 4.0 International}
|
222 |
+
}
|
223 |
+
```
|
224 |
+
|
225 |
+
<!--
|
226 |
+
## Glossary
|
227 |
+
|
228 |
+
*Clearly define terms in order to be accessible across audiences.*
|
229 |
+
-->
|
230 |
+
|
231 |
+
<!--
|
232 |
+
## Model Card Authors
|
233 |
+
|
234 |
+
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
|
235 |
+
-->
|
236 |
+
|
237 |
+
<!--
|
238 |
+
## Model Card Contact
|
239 |
+
|
240 |
+
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
|
241 |
-->
|